Performance verification of a confocal microscope for 3D metrology tasks

Dimensional and geometrical inspection of parts manufactured at micro and nanometric scales is currently difficult due to several limitations of possible inspection techniques. For example, SEM images provide only mostly qualitative information, while micro and nano precision CMMs require long measuring times and probe dimensions may limit the access to specific features. On the other hand, some optical microscopy techniques have the capability to acquire quantitative surface topography information as areal maps. Although they are primarily designed for surface topography measurements, they can be used to obtain geometric information as well. In particular, imaging confocal microscopy is a common technique for the 3D measurement of surface topography. Objectives used in confocal microscopy have a large numerical aperture that allows high lateral resolution (up to 0.2 μm) and high measurable local slope (up to 60°). Measurable local slope is a fundamental parameter when measuring parts such as MEMS, microfluidic channels, or optical devices. Beneath the capability of confocal microscopy to detect surfaces with high local slope, the measurement is affected by errors related to the slope angle of the surface with respect to the optical axis. This work proposes a method to easily quantify the magnitude of this error when measuring with a confocal microscope. The method is based on measurements of calibrated spheres and association of measurement errors to local surface slope.